The Universal vs. Targeted Debate: Should the United States Have Preschool for All?

Transcription

1 NATIONAL INSTITUTE FOR EARLY EDUCATION RESEARCH Contact Us: 120 Albany Street, Suite 500 New Brunswick, NJ What do we as a society want for our children? Few Americans would quarrel with the notion that all children should be ready to succeed in school and that no child should be left behind. Wide agreement on these broad goals reflects public awareness of research showing that learning is truly lifelong, beginning in the early years, and that early experiences build a foundation for learning. The Vision: Issue 6 / April 2004 The Universal vs. Targeted Debate: Should the United States Have Preschool for All? by W. Steven Barnett, Kirsty Brown and Rima Shore a shared agenda to ensure both a rewarding childhood and a promising future for all children. But how can these goals best be reached? And what is the role of government in pursuing them? On National Research Council 2000 these questions there is far less agreement. One key debate pits the notion of voluntary universal early learning programs, available to all preschoolers, against targeted services, reserved for those at greatest risk of poor achievement, based on economic disadvantage, disabilities or other special needs. Most public support for preschool programs today is for targeted programs but calls for universal programs have increased and several states seek to provide preschool for all 4-year-olds. 1 Tel (732) Fax (732) Website: nieer.org info@nieer.org What We Know: Targeted programs have lower costs, but do not realize other presumed advantages in practice. Universal programs are likely to be more effective at identifying and reaching all targeted children. School readiness is not just a problem of the poor. Young middle-income children lag behind their wealthy peers in social and cognitive skills. High-quality preschool has been found to benefit middle-income children, and added benefits could far exceed costs. Universal programs may have larger effects than targeted programs for the most disadvantaged children. Universal programs are likely to receive greater public support so that they are both of higher quality and reach more children than targeted programs. Policy Recommendations: The effectiveness and efficiency of investments in preschool could be increased with a shift from targeted to voluntary universal preschool programs. High quality standards for all children are required for effective universal preschool programs. Children with special needs due to poverty or disabilities may require more intensive services within universal programs. Expansion toward universal takes time, and patience is required to build capacity while maintaining or improving quality. Preschool programs could move toward universal access by gradually raising thresholds for eligibility. Federal matching funds could be used to encourage states to fund high-quality preschool for all.

2 The Case for Targeting Proponents of targeting cite three key advantages. Compared to universal preschool programs, targeted programs are said to be: (1) more efficient and cost the public less, (2) higher in quality, and (3) receive greater public support. Each of these arguments is set out more fully below. 2 Efficiency and Low Cost Targeted programs are said to have larger benefits and lower costs to the public. While high-quality early learning programs can benefit virtually all children, more substantial effects have been shown for those preschoolers most at risk of poor outcomes. Given this finding, why not invest resources where they are likely to do the most good? Moreover, targeted programs do not spend public dollars on children whose parents can afford such programs. Quality Because they serve a relatively small number of children with the greatest needs, targeted programs can focus on quality. They do not dilute quality by spreading resources too thin. These resources include not only money, but also facilities and qualified staff. Thus, targeted preschool programs are more likely to provide the intensity and duration of service required by children with the greatest needs. Public Support The smaller total budget required by a targeted program makes it more affordable and, therefore, more likely to be fully funded by the public. In addition, the public is more willing to pay for services when families cannot afford to purchase these on their own. Targeting is consistent with Americans historic preference for keeping most children in their mothers care, while providing out-of-home care for those whose home settings were considered inadequate. Our nation s first public preschools, the infant schools established in Massachusetts in the 1830 s, functioned on this principle, serving young children of the indigent and exposing them to mainstream values and habits. Public opinion continues to favor maternal care in an era when the great majority of mothers with young children are in the workforce for part or all of the day. 1 EC

3 The Growth of Targeted Programs Historically, policymakers across the nation have supported targeted programs based on the three apparent advantages enumerated above. Federal policy has supported a host of targeted programs, the largest of which are Head Start, preschool special education, and means-tested child care subsidies. In addition, all but 10 states fund preschool initiatives, and the vast majority target children deemed to be at risk of poor achievement by virtue of poverty or other factors. Most state programs do not charge a fee for these services. The exceptions are Massachusetts, Hawaii, Missouri, and Ohio where public preschool initiatives require payment by some parents. 2 The federal Head Start program dates back to Head Start was developed based on the idea that poverty severely restricted the capacity of many families and communities to adequately support the development of young children. Thus, Head Start was designed as a comprehensive child development program that works with families to improve children s health and nutritional status, social and emotional development, and cognitive development. Eligibility for Head Start is limited to young children in families with incomes below the federal poverty line or who potentially qualify for public assistance (TANF or SSI), with several exceptions, including allowing up to 10% of enrollment to qualify by virtue of a disability. Most Head Start participants are 3- and 4-year-olds, but it serves some children who are younger or older. Early Head Start is specifically designed to serve children under 3. In the year, Head Start s federal appropriation was nearly $6.7 billion. Federally funded enrollment in Head Start included more than 700,000 children at ages 3 and 4. Under the federal Individuals with Disabilities Education Act (IDEA) all states provide preschool special education to children with one of 13 disabilities or who meet state criteria for developmental delay. In , 357, and 4-year-old children were served in preschool programs under IDEA. Most of the cost is borne by state and local school districts. Data on these costs are not available, but we estimate total expenditure from all sources to be $5 billion in , based on a cost per child of $13, State-funded preschool programs expanded significantly over the last decade. By the school year, 40 states had preschool programs that enrolled 694,743 children. 4 These programs tend to serve only or mostly 4-year-olds. Although most programs target children from low-income families, Oklahoma and Georgia have implemented universal programs. Several other states are moving in the direction of universal preschool education: New York, West Virginia, and Florida. The last area of growth is public funding for child care services that often are provided in preschool classes, which can serve educational purposes. This growth reflects, in part, the growing number of children whose parents are not home during the day. The 2002 National Survey of America s Families showed that 82% of 3- and 4-year-olds with employed mothers are in nonparental care, and 43% use child care centers for this care. 5 Growth also reflects changes in welfare policies that require parents with young children to work or receive training away from home in order to qualify for public assistance. State-subsidized child care programs target children whose parents work or attend school and who meet a means test that is, family income must be below a state-established cutoff. In contrast to state preschool programs, most child care programs require parents to make co-payments that vary with income. 6 Growth in targeted preschool programs funded by the federal government and states suggests that, by some standards, the targeting movement has been successful. In 20 states over 25% of the 4-year-old population is served by Head Start, preschool special education, and state preschool. The vast majority of states serve 20% or more. In several states that have only targeted programs, over half the 4- year-olds are enrolled in these three programs. Including those served in state child care programs would push even more states over the 50% line (though the services offered by each type of program are not necessarily comparable). Curiously, similar success has not been extended to 3-year-olds, where only two states serve more than 25% and a total of seven enroll more than 20%. d At the same time, the case for universal preschool has gained momentum, winning the support of educational leaders most responsible for the achievement levels of American students, as well as business leaders, who arguably have the most to gain from raising those achievement levels. In some states, the push for universal may partially derive from the numerical success of the targeted programs. How targeted are these programs if they serve over half the population? Yet, the case also derives from serious concerns about how well targeted programs actually deliver on their promises regarding cost, efficiency, quality and public support. e 3

4 C The Case for Universal Preschool A C Proponents of universal preschool challenge all three 4 In recent years, some education and business leaders have concluded that universal preschool education is needed to improve school readiness and achievement. 7 The rationale often has been based on the concern that across the nation underachievement is a widespread problem, not limited to populations labeled at-risk. Achievement levels may be similar to those of previous generations, but today s children will face tougher challenges in a knowledge-based, global economy. Moreover, they say that our nation s competitors in the global marketplace are making investments in early education, and that failure to do so will dull our competitive edge. New insights into early development have shaped policymakers thinking as well. Given evidence that the foundation for literacy and other achievements is laid down in the early years, before children enter school, beginning universal educational services at age 5 or 6 seems arbitrary. The American public has long supported the right to a free public education. The question is: when should that right begin? In 1999, such reasoning persuaded the Council of Chief State School Officers representing the top state officials responsible for K-12 achievement to change a decade-old policy statement calling for preschool for at-risk 3- and 4- year-olds to a new policy calling for universal early learning programs. According to the Council, investments in K-12 education will not yield the results Americans want if children enter school without a strong foundation for learning. Similar concerns motivated the Committee for Economic Development (CED), a group representing corporate leaders, to make an unequivocal call for universal preschool. Noting the long-term social benefits of high-quality early learning programs, including lower arrest rates for participants when they reach the teen years, 8 law enforcement groups have supported universal preschool as well. 9 B presumed advantages of targeting. They argue that targeted programs: (1) are less efficient and cost the public more in the long run, (2) are lower in quality, and (3) receive less public support. Efficiency and Low Cost Many children who are not in targeted groups can benefit from a high-quality preschool education. The problems of low school readiness, low achievement, and dropout are not limited to the poor. And, targeted programs fail to reach many of the children they seek to serve. The costs of failing to serve children who could benefit are far higher than the costs saved by targeting. Quality Programs for the poor tend to be poor programs. Our cultural attitude toward charity programs is reflected in the proverb Beggars can t be choosers. Most targeted programs have not delivered the intensity or quality of educational services shown to be highly effective for children in poverty. Universal programs will tend to be of higher quality because they are not perceived as charity programs. In addition, universal programs may be more effective because they can serve disadvantaged children in more heterogeneous classes and all children benefit later when all of their schoolmates are better prepared for school. Public Support Although preschool for all will require a larger budget than targeted programs, it will nevertheless receive greater public support because of the larger, more influential population benefiting from the program. In addition, a universal program will be perceived as more fair and more in keeping with American s views that government has a responsibility to support education for all children. To some, the debate over universal vs. targeted preschool seems beside the point. Preschool education has become a necessity for middle-income families and many preschoolers have no parent at home during the day. Enrollment trends indicate the nation is quickly moving toward universal preschool whether we plan for it or not. d e

5 Preschool Policy Matters Figure 1. Social Skills of Entering Kindergarteners by Family Income Social Skills Score School Readiness Gap Bottom 20% 2nd Lowest Middle 20% 2nd Highest Top 20% 20% 20% Social Skills Optimal Development Family Income Comparing the Cases: What Does the Evidence Say? Given the almost diametrically opposed arguments in the debate over targeted vs. universal preschool programs, it is useful to consider the evidence and details of the arguments point by point. The growing momentum for universal preschool has prompted researchers to re-examine the theoretical advantages of targeted programs. What are the benefits for poor and non-poor children? How are targeted programs working in the real world of children and families? How well do universal programs work? Efficiency and Low Cost To what extent could all children benefit from effective preschools? In recent years, evidence has mounted that problems of school readiness and educational failure are not strictly problems of children in poverty. Many children arrive at school less than well prepared with respect to both social and academic skills that are important for school success. 10 Maryland, for example, found that only 52% of those who entered kindergarten in 2002 were fully ready. 11 In 2003, 37% of the fourth-grade population scored below basic on the reading section of the National Assessment of Educational Progress. Middle-income children have rates of grade repetition (12%) and high school dropout (11%) that are remarkably high, even compared to rates for children in poverty (17% and 23%). 12 Two diagrams display our analyses of the relationships between family resources and school readiness for children who entered kindergarten across America in Figure 1 shows the relationship between family income and an index of social skills that contribute to school success. 14 Figure 2 shows the relationship between family income and tests for reading ability (emergent literacy), math ability, and general knowledge. Both graphs show a nearly linear relationship with middle-income children far behind the richest 20%. Taking the performance of those children with the most resources as optimal, the gap between these children and everyone else represents the school readiness gap how far children from lower income families fall below the children of the top quintile. Clearly, children in poverty (contained in the bottom 20%) are far below others, but there is no clear cut-off where the gap dramatically declines and ceases to be important. There is substantial room for the vast majority of children to improve school readiness through better preschool education. 5

6 Comparing the Cases: What Does the Evidence Say? (continued) In addition, direct evidence has been growing that all children can benefit from high-quality preschool, including the more economically advantaged. While most experimental research has focused on disadvantaged children, at least one true preschool experiment has been conducted with more advantaged children. 15 This study found significant effects on achievement test scores in second and third grade for boys who attended preschool. Although effects were not significant for girls, they did score higher on 6 of 7 achievement subtests than girls who did not attend. Larger effects for middle-income boys than girls may be explained by boys higher rates of educational difficulties. Moreover, the small sample size was a clear limitation in trying to identify smaller effects for girls alone. Larger nonexperimental preschool studies yield additional evidence of preschool s benefits for the general population. In the United States, preschool has been found to increase early reading and math skills in kindergarten and first grade and decrease retention in kindergarten for all children. 16 Effects are somewhat larger for disadvantaged children, except for grade retention where rates for all children were reduced by about 25%. Strikingly similar results have been found by a national longitudinal study in Canada. 17 Research on the French preschool program suggests that each additional year of preschool improves achievement and that gains are similar for all socio-economic groups. 18 A large recent English study found positive effects of preschool on cognitive and social development. 19 Additional confirmation comes from substantial research literature indicating that the quality of child care influences the cognitive development of all children, with some finding stronger effects for disadvantaged children. 20 How imperfect is targeting? f h g children, and they may serve many children who are not Given the evidence that preschool s benefits are larger for some children, particularly the poorest, an argument can be made that targeted programs have the highest payoff. Of course, even if this is true, the smaller benefits for more advantaged children can still outweigh the costs. Putting this aside, targeted programs must actually serve the target population to attain this greater efficiency. In the real world, programs often fail to reach all of the target population and serve children who are not part of the target population. If these problems are serious enough, the gains from targeting can be lost and targeted programs can serve so few eligible children that they end up less efficient than a universal program. In fact, all of today s major targeted programs fail to reach a significant percentage of the eligible targeted. Forty years after its creation and nearly a decade after Congress authorized full funding, Head Start funding allows enrollment of less than 60% of 3- and 4-year-olds in poverty. Since so many poor children do not participate, and many children who participate are not poor (though they may be near-poor and meet the eligibility rules), Head Start undoubtedly reaches less than half the eligible population. Less than a third of children in poverty may be enrolled, far less than is generally assumed. Some of these children in poverty are enrolled in state-funded preschool programs, but how many is unknown, since no one knows how effectively states target and reach low-income children. Figure 2. Academic Abilities of Entering Kindergarteners by Family Income Academic Ability Scores School Readiness Gap Reading Math General Knowledge Optimal Development Bottom 20% 2nd Lowest Middle 20% 2nd Highest Top 20% 20% 20% Family Income Source: U.S.Department of Education, National Center for Education Statistics, Early Childhood Longitudinal Study, Kindergarten Class of , Fall 1998.

7 Turning to preschool special education, variations in enrollment across states raise serious questions about how well that population is identified. For example, the percentage of 4- year-olds in a state served by preschool special education programs varies from 3.7% to 12.5%. This is far greater than any likely variation in the target population. Moreover, the percentage of children enrolled in preschool special education at age 4 is on average about half the percentage of children enrolled in special education at ages 6-17 (9% to 16% by state). Thus, at least half the children eventually identified as needing special education are not identified for preschool special education services. Reaching the target population is an even more serious problem for low-income child care, primarily because of inadequate funding. Even assuming that all of the children served are eligible, only 12% of young children eligible for subsidies under the Child Care Development Fund block grant actually receive them. Many states have waiting lists. Budget cuts threaten to worsen the situation. Facing mounting deficits, many states with wait lists have refused new applications for child care subsidies. 21 The costs of targeting must be taken into account in evaluating the efficiency of targeted programs. It is costly and difficult to ensure that all eligible children are found and enrolled and all ineligible children are excluded. The eligible populations for targeted programs are constantly changing due to changes in the economy, family circumstances, and even children s abilities. It can be difficult even to estimate the size of the target population, much less identify all eligible children. No one has accurately estimated the size of the Head Start-eligible population. To effectively target, programs must find and enroll eligible children whose families do not apply to the program on their own. Reasons for not coming forward include not knowing that their children are eligible (e.g., parents may not know their child has a disability, or that their family income is below the poverty line) and not knowing what the eligibility guidelines are for the various programs. a The mobility of American families creates problems for targeted programs. Each year 22% of young children move to a new home, 14% within the same county and 8% to another county or state. 22 When children in Head Start, state preschool, or subsidized child care move they can find themselves without services because the program is either fully enrolled or not offered at their new location. As a result, many children receive less than a full year of services each year due to their families mobility. With eligibility for subsidized child care re-determined as often as monthly, state-administered child care programs have even greater turnover. In a five-state study, the average duration of continuous subsidy was only three to seven months, and even for those children who re-entered the system at a later time the continuity of their care suffered. 23 In California s part-day preschool program the total number of children enrolled at some time during was 135,000. The maximum number enrolled at any given time was 100,000. Thus, less than half of the children served (as few as 65,000) might have received a full preschool year. 24 Targeted programs for preschool children have no simple, low-cost way to find and screen the population for eligibility. Clearly, there is a problem with identification for preschool special education. Fewer than half the children who will need special education at some time are identified before kindergarten. When targeted programs miss many children who will later have problems, they fail to produce all of the possible benefits from prevention. A Closer Look at Targeting When it comes to getting services to the children and families who need them most, targeting may not be the best approach. Targeted programs: are inherently unfair: Some children qualify for good preschool programs free of charge, while other children from similar families get no services. may be just out of reach: Targeted programs often work against low-wage earners who play by the rules and whose hard-earned paychecks may keep their families just above the eligibility cutoff. cause disruptions as family status changes: Changes in a parent s work status, income, or home address can change eligibility. may not reach intended recipients: Some families who qualify for services will not enroll children, perhaps due to red tape involved in demonstrating eligibility or distaste for welfare programs; some who do not qualify will work the system to get their children in. e exclude middle-income children who could benefit: The effects may be smaller, but high-quality preschool benefits middle-income children, increasing achievement and preventing school failure. work against a coherent, efficient delivery system: Monies saved by serving only some children may be lost due to the fragmented delivery systems that result from multiple targeted programs. incur other hidden costs: Establishing and monitoring eligibility for millions of children can be very labor-intensive and costly. 7

8 Comparing the Cases: What Does the Evidence Say? (continued) What are the effects on costs? Most 3- and 4-year-olds in the U.S. are already in a classroom. The chart in Figure 3 shows for each region of the country in 2001: (1) the percentage of children participating in any type of early care and education and (2) the percentage of young children in an early care and education classroom. 25 To continue to treat preschool as an enrichment program for the disadvantaged would appear to be a form of denial one that misses opportunities to wisely invest in the children who are already attending preschool or child care and to ensure that others do not miss out altogether. As the CED asserted in calling for universal preschool, the nation now depends on a piecemeal and haphazard set of preschool arrangements that does not give all children equal opportunity to enter school ready to learn. 26 With so many children in preschool programs and child care, the nation is already spending significant public and private resources. Head Start alone spends nearly $7 billion per year. State-funded preschool adds another $2.5 billion in state dollars and an unknown amount of local funding. 27 Other sources of funding include federal and state child care subsidies and tax credits. Preschool special education may receive as much as $5 billion from the federal, state and local government. Nevertheless, the preschool programs most parents purchase now are relatively weak, and even publicly-funded targeted programs produce smaller effects than preschool programs of truly high quality. 28 The nation invests so little in the quality of these programs that we risk losing most of the benefits, which are many times the extra cost of a high-quality program. And, as Figure 3 shows, participation varies greatly depending on where in America a child lives. The CED has called for the government to spend at least $25 to $35 billion more to make good preschool programs available to all children at ages 3 and How much funding is needed will depend on how many families choose to participate and what is offered. Will government pay for a part day or a full day, for a school year or year round? Will parents continue to pay part of the cost? The CED estimate is certainly reasonable. Others have suggested as much as a $50 billion increase. Of course, all 4-year-olds could be enrolled for far less. To put these figures in perspective, $30 billion is less than one percent of total annual government spending, which exceeds $3 trillion, and a miniscule fraction of the domestic gross national product, projected to surpass $11.3 trillion in In sum, a universal high-quality preschool program would cost more and shift some costs from middle-income parents to taxpayers (many of whom are middle-income parents). Yet, the overall result could easily be large net gains to taxpayers. More complete coverage of high-need populations with high-quality programs could generate benefits that exceed the added costs of serving all children and there would be additional benefits from increased quality for middle-income children. Finally, the costs are well within the nation s means. Figure 3. Early Care and Education at 3 and 4; % 80% 60% 40% 20% 79% 74% 74% 66% 62% 57% 53% 46% Northeast South Midwest West 8 0% All Non-Parental Care 2001 Preschool Classroom Enrollment 2001

9 What About Quality? Two issues stand out regarding the quality of targeted and universal programs. First, to what extent are targeted programs of higher quality because they are able to focus resources on a smaller number of children? Second, might universal programs be more educationally effective because they include all of the children? Are targeted programs of high quality currently? If any program could demonstrate the benefits of targeting for quality, it should be Head Start. Yet, after 40 years, Head Start standards have never been raised to the levels of the preschool programs research has found to greatly improve the later school success of children in poverty. Head Start teachers are not required to have the qualifications expected of teachers generally and are paid only half the average salary of a public school teacher. Head Start spends less per child than the public schools even though smaller class sizes are required To some extent this is possible because state programs State preschool programs provide another source of evidence regarding the effects of targeting on quality. Most states target their programs, and all but two spend even less per child on their programs than the federal government spends on Head Start. About half the states have higher teacher qualifications requirements than Head Start. provide less intensive 2 services outside the classroom, but it also reflects contributions from the local schools and other sources to cover a portion of costs. The two universal state programs have relatively high standards, though not the highest. 31 Preschool special education programs are by far the best funded per child. Federal law requires states to provide children with disabilities a free appropriate education beginning at age 3. However, the federal government caps its spending for the program, and federal preschool special education funding has steadily declined for many years on a per-child basis. 32 States and local school districts have had to bear the vast majority of the costs of this program. One lesson from this experience might be that adequate funding depends on strong legal entitlements enforced by the courts. In this regard, it is noteworthy that the best funded state preschool program for disadvantaged children is in New Jersey, where the state Supreme Court intervened to require high-quality preschool education The quality of state-subsidized child care is poor to mediocre on average and highly variable. 34 Child care standards are considerably lower than standards for most state preschool programs. Increases in public funding for child care over the last decade have sought primarily to facilitate workforce participation by low-income mothers those most likely to receive public assistance. This focus led to a significant boost in federal and state funding for subsidies to lowincome parents, with the funds used to increase the number of slots available at the expense of quality. According to policy analysts Richard Brandon, Sharon Lynn Kagan, and Jutta Joesch, it resulted in many states trying to buy the greatest quantity of care for the lowest cost. They add that U.S. social welfare policy has tended to provide benefits designated for the poor at minimum levels without regard for quality. 35 Although children from higher-income families generally have more access to programs than do poor children, the quality of the programs they attend is not necessarily high. The quality of education available to young children today depends on their family s income, where they live, whether they qualify for various targeted programs, and parental employment. 36 Replacing today s haphazard, fragmented services with a universal program would allow states to strengthen quality by setting and monitoring minimum standards that would apply to all programs, regardless of their auspices (public school, for-profit, nonprofit). In a system where everyone receives the same quality, the state might ensure a higher level of quality than when only the poor receive services. Such a system is likely to increase public dialogue about the benefits of preschool and what high quality looks like. 9

10 Might universality improve the quality and effectiveness of preschool programs? 10 Several reasons have been advanced for the claim that universal preschool programs might be of higher quality. The most basic of these is that they would no longer be viewed as charity programs, and a broader and more influential cross-section of the nation would have a direct stake in their quality. Middle-income parents are more organized and influential in advocating for quality preschool special education than for other preschool programs. If included in a universal system, they would advocate for quality in that system. Finally, universal preschool programs have the potential to produce significantly greater long-term benefits because everyone participates in the program. Most studies of the effects of high-quality preschool programs are based on small experiments or targeted programs that do not serve most children, even in communities with high concentrations of disadvantaged children. Such studies fail to capture the additional benefits that may accrue when all children participate. For example, if all of the children in an elemen- 1tary 3 school had attended a high-quality preschool program, 2 A universal program might strengthen public commitment to quality, while providing a legal foundation for insisting on quality for all children. Consumers of early childhood education services may become better informed and better able to define and demand high-quality programs, thus maintaining higher standards among public and private providers. In addition, universality may directly increase the educational effectiveness of preschool education. A universal approach would remove the stigma associated with virtually all targeted programs, while lowering the barriers to the integration of children from diverse ethnic, cultural, and socioeconomic backgrounds, as well as those with and without disabilities. 37 Stigma may not only discourage parents from enrolling their children, but it may also cause program staff to treat children and parents in ways that would be unacceptable in a universal program. Also, interactions with a broader range of peers may improve educational outcomes for children with disabilities and other disadvantages while educating all children in their attitudes and behavior towards others. 38 One recent small-scale study found that low-income children in economically-integrated preschool programs had much larger language gains than their counterparts in programs that served only low-income children. 39 the overall academic and social climate would change. This could further boost learning and development. Once the effects of preschool permeated a neighborhood, city, or state the overall quality of life would improve less school failure, less crime and delinquency, higher employment rates for young adults, and increased earnings. Such changes would reinforce the direct effects on children. Even relatively small pervasive changes could produce much larger benefits if they push schools and communities past their tipping points where the rates of school behavior problems or crime rapidly decline. 40 4

11 Preschool Policy Matters The Term Universal Is Not Universal It means different things to different people and public support for universal public preschool likely depends on how it is defined. Here are some ways that universal preschool has been interpreted. Universal = Free and available to all. Like K-12 education, preschool is free to anyone who wants to participate. This could mean requiring that school districts offer public education for one or two years before kindergarten entry. Some state programs offer funds to all school districts, but do not require all districts to participate. Universal = Affordable and available to all. Programs are available for all children whose parents want them to attend. While programs may not be completely free, they are affordable because some part of the program is free and/or fees are related to family incomes. Universal = Compulsory. Since public kindergarten has existed for a century and is mandatory only in a few states, this seems highly unlikely. Voluntary enrollment of young children in educational programs is a value held strongly by the public and policymakers. Universal = Guaranteed subsidy. Finance mechanisms (vouchers, tax credits, loans) are offered to all families, but no agency is responsible for ensuring that quality programs are available. For more discussion, see A.W. Mitchell. April Education for All Young Children: The Role of States and the Federal Government in Promoting Prekindergarten and Kindergarten. New York: Foundation for Child Development. 5 11

12 What About Public Support? 12 The question of whether targeted programs garner greater public support because they cost less and focus resources on those with least ability to pay is not limited to preschool education. Political scientist Theda Skocpol has studied the issue generally and finds that programs designated for certain groups have trouble maintaining political support over the long term, as other interest groups vie for government funding. The most successful measures have been those that ensured entitlements to broad categories of beneficiaries, says Skocpol. There has been no political backlash against social security. Even in a generally conservative period it has been protected by its broad constituency; and it has continued to be championed by congressional representatives of all partisan and ideological stripes. 41 By contrast the War on Poverty quickly generated political backlashes that ended possibilities for its continuation and improvement. During the 1970s, public opinion polls recorded decreasing support for government efforts to aid minorities and for public social spending, especially on service programs popularly identified with poor blacks. 42 Economists Jonah Gelbach and Lant Pritchett have applied economic theory to this issue and find that despite the lower costs of targeted programs, universal programs can be sound economic policy and maximize well-being of society as a whole. 43 They conclude that lack of political support for means-tested programs when budgets are determined by majority voting can lead to such small budgets for meanstested programs that even the poor and middle classes are worse-off with means-tested than with universal programs. 44 Their theoretical conclusions are consistent with the evidence on targeted preschool programs discussed earlier. History provides abundant evidence of weak political support for targeted preschool programs. In four decades, Head Start has failed to achieve full funding for a quality program. The political practicality and affordability of universal preschool programs has had much less time to be tested. Two states can plausibly claim to have universal preschool for 4-year-olds: Georgia and Oklahoma. In both, the program is relatively new. Georgia switched from targeted to universal preschool at least in part because of fears that the targeted program would not survive politically. Georgia s universal program achieved a high degree of popularity. However, the program was changed in ways that suited middle-income families better (e.g., discontinuing family support services) and state spending per child declined after the program was converted. 45 Quality standards for Georgia s program remain higher than average for a state program. Oklahoma s universal program serves a slightly higher percentage of children than Georgia s. In Oklahoma, the program is essentially part of public education and operates with public school standards. Teachers have college degrees with specialized training in early childhood education. In Oklahoma, as in Georgia, the percentage of the population served has increased each year, though it remains short of 100% participation. Experience in another state seeking to implement a universal preschool program has been more mixed. In New York, growth of universal preschool has been at least temporarily stymied by budget maneuvers that limit the program largely to disadvantaged children. However, efforts to eliminate the program entirely have been repeatedly defeated, and the number of children served is much larger than under the previous targeted program alone. Two other states provide only an indication that universal preschool education can have a broad appeal. In 2003, West Virginia adopted a plan to provide universal preschool for 4-year olds by the school year. A year earlier, Florida voters approved an amendment to the state constitution requiring high quality preschool education for all 4 year-olds by Voters approved this initiative directly after the Legislature failed to pass legislation for this purpose. However, initial legislative proposals to implement the Florida amendment fall far short of high quality. d B e

13 Policy Conclusions and Recommendations Voluntary, universal preschool programs providing access to high-quality preschool education to all children may be more educationally effective and economically efficient than targeted programs. Targeted programs have lower costs and avoid shifting costs to the taxpayers for families who can already afford preschool programs. However, other presumed advantages of targeted programs often are not realized in practice. Determining eligibility for targeted programs is costly, difficult, and imperfect. Many eligible children may be overlooked while ineligible children receive services. Frequent changes in eligibility and mobility of families pose problems. Universal programs may be more effective at ensuring that all children targeted by current programs are actually identified and served. No clear divide separates disadvantaged children who can benefit from preschool education from other children. Instead, there is a continuous relationship between income and social and cognitive skills at kindergarten entry, and preschool education has been found to benefit middleand higher-income children. Universal programs may be more effective for disadvantaged children because they serve all children. As some research shows, classrooms serving middle-income and poor children together are more effective. Despite higher costs, universal programs may obtain stronger public support for adequate funding and high quality. d Given the practical difficulties of creating new programs and the higher rate of current enrollment of 4-year-olds than 3-year-olds, the first step toward better policy for many states may be to create universal programs at age 4. Several states have already taken this step. Expansion toward universal coverage takes time. States should be careful not to move so fast toward universal coverage that they seriously compromise quality. Many children already have access to low- or mediocre-quality preschool programs that squander potential benefits. Any approach that did not first serve children currently eligible under a targeted program would not be politically feasible. States can move toward a universal coverage gradually for example, by raising the income threshold over time or offering the universal program first in high need communities. Universal programs should have high minimum quality standards for all children, but recognize that one size fits all is poor educational policy. Supplemental services can be offered to children with greater needs targeting within universal. This is particularly important for families who need help meeting children s nutrition and health needs. High-quality preschool for all 3- and 4-year-olds would require roughly $30 billion in additional public funding. This is only one percent of total government spending. The added cost to serve all 4-year-olds could be far less. 13

14 Endnotes: 1 Casper, L.M. & Bianchi, S.M. (2002). Continuity and change in the American family. Thousand Oaks, CA: Sage Publications, Inc. 2 Barnett, W. S., Robin, K.B., Hustedt, J.T., & Schulman, K.L. (2003). The state of preschool: 2003 state preschool yearbook. New Brunswick, NJ: NIEER. 3 Chambers, J., Shkolnik, J., & Perez, M. (2003). Total expenditures for students with disabilities, : Spending variation by disability. Washington, DC: US Department of Education, Office of Special Education Programs, American Institutes for Research. 4 Barnett, W.S., Robin, K.B., Hustedt. J.T., & Schulman, K.L. (2003). 5 Capizzano, J. & Adams, G. (November, 2003). Children in low-income families are less likely to be in center-based child care. Snapshots 3 of America s families (No. 16). Washington, DC: Urban Institute. 6 It is difficult to determine with any precision the level of government spending and numbers of 3- and 4-year-olds served for child care programs. 7 Committee for Economic Development (2002). Preschool for all: Investing in a productive and just society. Washington, DC: Committee for Economic Development. National Association of Elementary School Principals (2004). Early education ensures that we leave no child behind: A principal s perspective. 8 Reynolds, A. J., Temple, J.A., Robertson, D.L., & Mann, E.A. (2001). Long-term effects of an early childhood intervention on educational achievement and juvenile arrest: A 15-year follow-up of low-income children in public schools. JAMA, 285 (18), Fight Crime Invest in Kids (2004). Quality pre-kindergarten: Key to crime prevention and school success. Washington, DC: Fight Crime: Invest in Kids. Available at 10 West, J., Denton, K., & Germino-Hausken, E. (2000). America s kindergarteners: Findings from the Early Childhood Longitudinal Study, kindergarten class of , Fall Washington, DC: National Center for Education Statistics, U.S. Department of Education. 11 Bowler, M. (2003). Fifty-two percent of kindergarteners in Maryland judged fully ready. Baltimore Sun. 12 Multi-year averages based on figures from: National Center for Education Statistics. (1997). Dropout rates in the United States: Washington, DC: U.S. Department of Education. 13 Data are from the Early Childhood Longitudinal Study Kindergarten Cohort (Fall 1998) conducted for the National Center for Education Statistics. 14 The index was constructed by summing across 3 different social outcome measures including approaches to learning, self-control and interpersonal social ratings. 15 Larsen, J., & Robinson, C. (1989). Later effects of preschool on low-risk children. Early Childhood Research Quarterly, 4, Magnuson, K., Meyers, M., Ruhm, C., & Waldfogel, J. (2003). Inequality in preschool education and school readiness. New York : Columbia University. 17 Lipps, G., & Yipton-Avila, J. (1999). From home to school: How Canadian children cope. (Culture, Tourism and the Center for Education Statistics, Catalogue # 89F0117XIE). Ottawa, Ontario: Statistics Canada. 18 Hirsch Jr., E.D. (2004). Equity effects of very early schooling in France. Core Knowledge Preschool. Available: 19 Sammons, P., Sylva, K., Melhuish, E., Siraj-Blatchford, I., Taggart, B., & Elliot, K. (2002). Measuring the impact of pre-school on children s cognitive progress over the preschool period (Technical paper 8a). London: Institute of Education, University of London. Sammons, P., Sylva, K., Melhuish, E., Siraj-Blatchford, I., Taggart, B., & Elliot, K. (2003). Measuring the impact of pre-school on children s social/behavioural development over the preschool period (Technical paper 8b). London: Institute of Education, University of London. 20 Lamb, M. (1996). Effects of nonparental child care on child development: An update. Canadian Journal of Psychiatry, 41, Burchinal, M. R., Roberts, J. E., Riggins, R., Zeisel, S. A., Neebe, E., & Bryant, D. (2000). Relating quality of center child care to early cognitive and language development longitudinally. Child Development, 71, NICHD Early Childcare Research Network (2002). Child-care structure, process, outcome: Direct and indirect effects of child care quality on young children s development. Psychological Science, 13 (3),

15 21 Greenberg, M., Mezey, J., & Schumacher, R. (March 2003). Child Care Funding: The Story Since 1996 and The Challenges in Reauthorization. Washington, DC: Center for Law and Social Policy. 22 Hodgkinson, H. (2003). Leaving too many children behind: A demographer s view on the neglect of America s youngest children. Washington, DC: Institute for Educational Leadership. 23 Meyers, M., Peck, L., Davis, E., Collins, A., Kreader, J.L., Georges, A., Weaver, R., Schexnayer, D., Schroeder, D., & Olson, J. (2002). The dynamics of child care subsidy use: A collaborative study of five states. New York: National Center for Children in Poverty. 24 California Department of Education (2003). CD800, Child care annual aggregate report, FFY Sacramento, CA: Child, Youth, and Family Services Branch, California Department of Education. 25 Barnett, W.S., Yarosz, D., & Thomas, J. (2004). Where are the kids? New Brunswick: NIEER. 26 Committee for Economic Development (2002). 27 Barnett, W. S., Robin, K.B., Hustedt, J.T., & Schulman, K. L. (2003). 28 Barnett, W.S. (1998). Long-term effects on cognitive development and school success. In W.S. Barnett and S.S. Boocock (Eds.) Early care and education for children in poverty: promises, programs, and long-term outcomes, (pp.11-44). Buffalo, NY: SUNY Press. Magnuson, K., Meyers, M., & Ruhm, et al., (2003). 29 Committee for Economic Development (2002). 30 Barnett, W. S. (2003). Investing in Head Start teachers. Preschool Policy Matters (4). New Brunswick, NJ: NIEER. Barnett, W. S., Robin, K.B., Hustedt, J.T., & Schulman, K. L. (2003). 31 Barnett, W. S., Robin, K.B., Hustedt, J., & Schulman, K. L. (2003). 32 National Center for Educational Statistics (2002). Digest of education statistics, Washington, DC: U. S. Department of Education. 33 Barnett, W.S., Robin, K.B., Hustedt. J.T., & Schulman, K.L. (2003). 34 Cost, Quality and Outcomes Study Team. (1995). Cost, quality, and child outcomes in child care centers. Denver, CO: Economics Dept., University of Colorado at Denver. 35 Brandon, R.N., Kagan, S.L., & Joesch, J.M. (2000). Design choices: Universal financing for early care and education. Financing Universal Early Care and Education for America s Children Project. Available at designchoices.pdf. 36 Barnett, W.S., Robin, K.B., Hustedt, J.T., & Schulman, K.L. (2003). Fuller, B., & Strath, A. (2001). The child-care and preschool workforce: Demographics, earnings, and unequal distribution. Educational Evaluation and Policy Analysis, 23 (1), Phillips, D., Mekos, D., Scarr S., McCartney, K., & Abbott-Shim, M. (2000). Within and beyond the classroom door: Assessing quality in child care centers. Early Childhood Research Quarterly, 15 (4), Garfinkel, I. (1996). Economic security for children: From means testing and bifurcation to universality. In I. Garfinkel (Ed.) Social policies for children (pp ). Washington, DC: Brookings Publications. 38 Odom, S. L. (2000). Preschool inclusion: What we know and where we go from here. Topics in early childhood special education, 20 (1), Schecter, C. (2002). Language growth in low-income children in economically integrated versus segregated preschool programs. West Hartford, CT: St. Joseph College. 40 Gladwell, M. (2000). The tipping point: How little things can make a big difference. New York: Little, Brown, & Company. 41 Skocpol, T. (1991). Targeting within universalism: Politically viable policies to combat poverty in the United States. In P.E. Peterson & C. Jencks (Eds.) The urban underclass. Washington, DC: Brookings Institution. 42 Skocpol, T. (1991). 43 Gelbach, J., & Pritchett, L. (2002). Is more for the poor less for the poor? The politics of means-tested targeting. Topics in economic analysis and policy, 2 (1), Article 26. Avaiable at: 44 Gelbach, J., & Pritchett, L. (2002). 45 Raden, A. (1999). Universal prekindergarten in Georgia: A case study of Georgia s lottery-funded pre-k program. Working paper. New York: Foundation for Child Development. 15

16 by W. Steven Barnett, Ph.D., Kirsty Brown, M. Ed, and Rima Shore, Ph.D. W. Steven Barnett is director of the National Institute for Early Education Research. Dr. Barnett s research has focused on the long-term effects of preschool programs on children s learning and development, the educational opportunities and experiences of young children in low-income urban areas, and benefit-cost analyses of preschool programs and their long-term effects. He received his Ph.D. in economics from the University of Michigan. He is a Professor of Education Economics and Public Policy at Rutgers University. Kirsty Brown is assistant director of the National Institute for Early Education Research. Rima Shore holds the Adelaide Weismann Chair in Educational Leadership at Bank Street College of Education. The Targeted vs. Universal Debate: Should the United States Have Preschool for All? is issue 6 in a series of briefs, Preschool Policy Matters, developed by The National Institute for Early Education Research. It may be reprinted with permission, provided there are no changes in the content. Available online under Resources/NIEER Publications at nieer.org This document was prepared with the support of The Pew Charitable Trusts. The Trusts Starting Early, Starting Strong initiative seeks to advance high quality prekindergarten for all the nation s three-and four-year-olds through objective, policy-focused research, state public education campaigns and national outreach. The opinions expressed in this report are those of the authors and do not necessarily reflect the views of The Pew Charitable Trusts. NATIONAL INSTITUTE FOR EARLY EDUCATION RESEARCH 120 Albany Street, Suite 500 New Brunswick, New Jersey (Tel) (Fax) Website: nieer.org Information: info@nieer.org